Expert Financial Analysis and Reporting

I put out a blog last week alerting subscribers that I am writing an update on Northwest Biotherapeutics. I continue to work on this report and hope to have it out soon, but I am compelled to address some late breaking news beforehand. As the report stood yesterday, I was highlighting as a major uncertainty one aspect of the Statistical Analysis Plan (SAP). Northwest has specified in the SAP that it will compare results of DCVax-L plus standard of care (SOC) from the phase 3 trial in newly diagnosed glioblastoma ndGBM to SOC results from other phase 3 trials in ndGBM that were conducted in the same time frame as the DCVax-L trial. This use of external controls was unthinkable some years ago as statistical purity demanded that  patients enrolled in a trial be randomized so that some patients were randomized to active drug added to SOC and others to just SOC. Cross trial comparisons were unthinkable.

This aversion to use of external control has been eroding in recent years as the FDA began to consider that use could speed development of breakthrough drugs. As an example, the CAR-T drugs were initially approved on the basis of phase 2 trials that had no control arm and essentially relied on external controls. However, the FDA had not taken a formal position on external controls. Short sellers have stridently maintained that the FDA would not even consider the phase 3 trial of DCVax-L because it uses external controls so that the trial was doomed.

Based on the example of the CAR-T drugs, I was reasonably certain that the FDA would accept external controls. My confidence was also buoyed by action by the MHRA, the UK equivalent of the FDA. On its website, it indicated that it would accept NWBO's SAP use of external controls. Still, there was no clear signal from the FDA on whether it agreed with the MHRA decision. Yesterday, it issued a statement that clarifies its position on the use of external control arms in oncology trials. I include the statement from FDA at the tail end of this note. It makes clear that for the first time that I am aware of that the FDA will consider and probably allow the use external controls to evaluate the DCVax-L trial. It was signed by 13 FDA officials including Richard Pazdur. He is key player at the FDA as shown by a brief FDA bio which is also shown in the back of this note.

This is an extremely encouraging development for Northwest that comes on the heels of an equally exciting announcement that the MHRA had GMP certified the Company’s Sawston, UK plant for production of DCVax-L for compassionate use. It is clear from the FDA statement that any company using external controls in a regulatory submission will have to rigorously address issues about the external controls. In the case of DCVax-L, Northwest will have to address numerous issues in the external control arm such as methylation status, IDH mutations, degree of surgical resection and so on. It will have to convince the FDA that patients in the external control arm are medically comparable to patients treated with DCVax-L plus SOC in the phase 3 trial. This is a difficult challenge for NWBO and I suspect that gathering this information has been a major factor for the Company not yet releasing topline data from the phase 3 trial in which the data lock occurred on October 5, 2020.

FDA Statement on External Control Arms in Oncology: Current Use and Future Directions

• S. Mishra-Kalyani
• Amiri Kordestani
• R. Rivera
• Singh
• Ibrahim
• A. DeClaro
• Shen
• Tang
• Sridhara
• G. Kluetz
• Concato
• Pazdur
• A. Beaver

If tumor response cannot be reasonably measured due to disease characteristics (such as in certain neuro-oncologic tumors) or there is interest in estimating a comparative treatment effect within the population of interest, approaches have been explored to supplement single arm data with data external to the clinical trial, also referred to as an external control arm.

A fit-for-purpose assessment is necessary as data source quality and granularity varies. Best practices for study design and statistical analysis plans are not yet defined and challenges exist. Despite these challenges, marketing applications in oncology have included trial designs with external control comparator arm(s). The adequacy of the external control data has been variable based on the quality and it being fit-for-purpose in a comparative analysis.

Ventz et al considered the use of data from both previously conducted clinical trials as well as RWD for an externally controlled trial to understand treatment effect on OS in glioblastoma multiforme (GBM). The authors advocate for the use of several sources of external data to assess and potentially reduce potential bias, provide complete time-to-event data, and retain sufficient power when comparing to traditional randomized control trial designs.

Additionally, the 21st Century Cures Act has led FDA to develop a framework for evaluating the quality and relevance, also referred to as fit-for-purpose, of real-world evidence (RWE), such that it may be incorporated into regulatory decision making.

Potential applications may include a study design for a clinical question where the natural history, morbidity, or mortality of the disease is well characterized, highly predictable, the expected effect size of the investigational treatment is high, and the outcome precisely measured.

It is important that selection of data eligible for an external control arm is transparent and traceable with an audit trail available for FDA inspection.

A pre-specified statistical analysis plan for any study that incorporates external control data increases the integrity of the data analysis and results. The analysis plan should include statistical methods to account for various types of potential bias, including major threats from lack of randomization (e.g., selection bias) and confounding amongst others. An assessment of the similarity of the patient populations in each arm using pre-specified criteria to measure balance [16, 17], before and after any statistical procedures or adjustments could further minimize the concern of bias. These comparisons of the populations of interest before and after analytical adjustments to account for bias and confounding are ideally conducted prior to the analysis of any outcome data.

To date, no primary efficacy analysis of a study used to support approval of a marketing application in oncology has included a formal comparison to an external control arm.

In the future, it is possible that formal analyses for comparative treatment effect between an investigational therapy and an external control arm would provide the primary evidence of efficacy to support a regulatory approval in oncology, particularly for a disease type with a highly predictable natural history and the precisely measurable treatment effect.

Importantly, the granularity of the data with respect to measurement or categorization of important variables is not trivial, and must be similar between experimental arm and external control data to ensure comparative analyses.

Bio for FDA’s Richard Pazdur

Richard Pazdur, M.D. is the director of the FDA's Oncology Center of Excellence (OCE), which leverages the combined skills of the FDA's regulatory scientists and reviewers with expertise in drugs, biologics and devices to expedite the development of novel cancer products. In his role as director of the OCE, Pazdur is responsible for leading the effort to develop and execute an integrated regulatory approach to enhance the cross-center coordination of oncology product clinical review.

Pazdur previously served as the director of the Office of Hematology and Oncology Products (OHOP) in the FDA’s Center for Drug Evaluation and Research and will continue to serve in OHOP as acting director. This Office was formed in 2005 to consolidate the review of drugs and therapeutic biologics for the diagnosis, treatment, and prevention of cancer, as well as the review of drugs and therapeutic biologics for hematologic diseases and for medical imaging. As director of OHOP, Pazdur facilitated coordination of oncology activities across all FDA Centers and ensured an ongoing outreach and collaboration between the FDA, the National Cancer Institute, and other cancer-related organizations within and outside of the government. Pazdur was the director of the Division of Oncology Drug Products from September 1999 to May 2005.